The cloud computing environment is an enhancement to the predecessor grid environment, whereby multiple grids and other computation resources may be further abstracted by a cloud layer, thus making disparate devices appear to an end-consumer as a single pool of seamless resources. These resources may include such things as physical or logical compute engines, servers and devices, device memory, storage devices.
Managing performance oriented SLAs of cloud based operations is a relatively new problem has many technical and practical challenges. These challenges include ensuring customer satisfaction and avoidance unnecessary financial and economic costs. As an example, over-delivery of services typically equates to higher cost and resource usage than is required, while under-delivery often works to the detriment of the service provider. Typical solutions monitor for drop-offs in performance, but typically do not take proactive measures to prevent failure in meeting SLA terms.